University of Twente Student Theses
Improving speech emotion recognition by identifying the speaker with Multi-task Learning
Schrama, M.L. (2020) Improving speech emotion recognition by identifying the speaker with Multi-task Learning.
|
PDF
1MB |
| Abstract: | Automatic emotion recognition is an important topic in artificial intelligence as it can improve our experience when interacting with machines. This paper proposes a multitask learning (MTL) algorithm for speech emotion recognition as the main task and speaker identification as the auxiliary task. After experimenting both separate training and MTL training on the same dataset, the proposed method results in a relative improvement of 15.6% on the accuracy of emotion recognition. Additionally, it is observed that the MTL model does not make a contribution on the correctness of identifying speakers. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/82143 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page